Quickstart

The whole path, end to end. Each step is self-contained — if you already have an account and a Disk, start at step three.

1. 1

   Create a free account

   Sign up at skilldiscs.com/register. No credit card — the free tier includes API access at 100 requests per minute.

2. 2

   Add a Disk

   Paste a URL, upload a PDF, drop in a YouTube link, or connect a GitHub repository. SkillDiscs parses it into sections, extracts key points, and generates embeddings. This is the step that replaces your own parsing and chunking pipeline.

3. 3

   Generate an API key

   Open Settings → API Keys and create a key. Keep it server-side: any holder can read every Disk in your library, including private ones.

4. 4

   Send a semantic search request

   POST to https://skilldiscs.com/api/v1/search with an Authorization: Bearer header and a JSON body containing your query. You get back matching sections with similarity scores, key points, and source attribution.

5. 5

   Connect the result to your agent

   Feed the returned sections straight into your agent as retrieval context, point an MCP client at the endpoint, or export the Disk as a Claude Skill. The content is already chunked and cited, so nothing further is required on your side.

SkillDiscs exposes a small REST API at https://skilldiscs.com/api/v1. Three endpoints cover the full retrieval surface:

• POST /search — semantic search across all your Disks (returns matching sections with similarity scores)
• GET /disks/:id — full structured content of a single Disk
• GET /disks — list / filter your Disks

Every section is returned with key points, embeddings, and source attribution — already chunked for RAG. No HTML parsing, no PDF extraction, no chunking heuristics on your side.

All requests use Authorization: Bearer <api_key>. Generate a key from your Settings → API Keys page after you sign in.

Authorization: Bearer sk_live_abc123…
Content-Type: application/json

The most common call: ask a natural-language question, get the most relevant Disk sections back, sorted by cosine similarity. Use this directly as RAG context for your model.

curl -X POST https://skilldiscs.com/api/v1/search \
  -H "Authorization: Bearer $SKILLDISCS_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "how does photosynthesis work?",
    "limit": 5
  }'

Example response

{
  "results": [
    {
      "disk_id": "k7x2m9nq4b",
      "disk_title": "Plant Biology Fundamentals",
      "section_index": 4,
      "section_text": "Photosynthesis converts light energy …",
      "key_points": [
        "Chloroplasts capture sunlight",
        "Calvin cycle fixes CO₂"
      ],
      "similarity": 0.94,
      "source_url": "https://example.com/article"
    }
  ]
}

Fetch the full structured content of one Disk by id.

curl https://skilldiscs.com/api/v1/disks/k7x2m9nq4b \
  -H "Authorization: Bearer $SKILLDISCS_KEY"

Browse / filter your library. Supports ?category, ?language, ?limit, ?cursor.

curl "https://skilldiscs.com/api/v1/disks?category=science&limit=20" \
  -H "Authorization: Bearer $SKILLDISCS_KEY"

Default: 100 requests / minute per API key. Exceed it and you'll get a 429 with a retry_after field. Need more? Email us — we lift it for paid plans on request.